# High Robustness

xFinder-llama38it is a fine-tuned key answer extraction model based on Llama3-8B-Instruct, designed to improve the accuracy and robustness of key answer extraction from large language model outputs.

xFinder-qwen1505 is a model specifically designed for key answer extraction in large language models (LLMs), fine-tuned from Qwen-1.5-0.5B.

This model is released under the AFL-3.0 license; specific details are currently unavailable.

A Nynorsk automatic speech recognition model based on the XLSR-300M architecture, trained on the NPSC dataset with low word error rate and character error rate.